Attribute reduction in decision-theoretic rough set models using genetic algorithm

  • Authors:

  • Srilatha Chebrolu;Sriram G. Sanjeevi

  • Affiliations:

  • Department of Computer Science and Engineering, NIT Warangal, AP, India;Department of Computer Science and Engineering, NIT Warangal, AP, India

  • Venue:
  • SEMCCO'11 Proceedings of the Second international conference on Swarm, Evolutionary, and Memetic Computing - Volume Part I
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Real world data may contain inconsistencies, uncertainty and noise. Rough set model is a mathematical methodology in data analysis to deal with inconsistent and imperfect knowledge. Various probabilistic approaches to rough set model are proposed. Decision-theoretic rough set model (DTRSM) is one of the probabilistic approaches to rough set model. This paper proposes an attribute reduction algorithm in DTRSM, through region preservation. Attribute reduction is the process of identifying and removing redundant and irrelevant attributes from huge data sets, reducing its volume. The reduced data set can be much more effectively analyzed. Attribute reduction in DTRSM through region preservation is an optimization problem, thus Genetic Algorithm (GA) is used to achieve this optimization. Experiment results on discrete data sets are compared with local optimization approach based on discernibility matrix method and has been shown that GA can be effectively and efficiently used to achieve global minimal reduct.